My problem can be simplified to the following example:
create table dsort(dfrom date,dto date,pname varchar(10));
insert into dsort(dfrom,dto,pname) values(20180101,20180101,'Anja');
insert into dsort(dfrom,dto,pname) values(20190101,20191231,'Fred');
insert into dsort(dfrom,dto,pname) values(20190201,20190201,'Willy');
insert into dsort(dfrom,dto,pname) values(20190301,20190301,'John');
insert into dsort(dfrom,dto,pname) values(20191230,20200131,'Sepp');
insert into dsort(dfrom,dto,pname) values(20200201,20200202,'Leo');
Now I'm looking for an sql statement which selects all rows (ordered by dfrom
) having a dfrom
not between the dfrom
and dto
of any previously selected row, possibly using window functions to get informations on other rows, like this:
select dfrom,dto,pname from ...
resulting in:
dfrom dto pname
20180101 20180101 'Anja'
20190101 20191231 'Fred'
20200201 20200202 'Leo'
Can anyone give me the necessary sql statement?
I think the simplest approach is not exists
:
select d.*
from dsort d
where not exists (
select 1
from dsort d1
where d.dfrom > d1.dfrom and d.dfrom < d1.dto
)